VMem-Bench (dataset)

Textual gold and SUT prompts for VMem-Bench. No source videos.

Split	Contents
trackA/	Per-movie gold/{entity_registry,chunk_annotations,chunk_index}.json (Blender Open Movies + LSMDC titles)
trackB/	50 long-horizon stories: gt, gt_source, sut_prompts (zh/en) + freeze_manifest.json

• Code / scoring: https://github.com/Suchenl/VMem-Bench
• Method: https://github.com/Suchenl/MemStrata
• License of these JSON files: CC BY 4.0 (LICENSE)
• LSMDC videos: apply at the LSMDC download page, stitch to LSMDC_Videos_Stitched/<movie_id>.mp4; cite Rohrbach et al., IJCV 2017. Full steps: VMem-Bench docs/DATA.md
• Blender videos: official project pages / scripts/prepare_blender.sh for BBB; same DATA.md

See trackA/README.md and trackB/README.md in this dataset for file schemas.

Citation

@article{chen2026memstrata,
  title={Stratifying and Benchmarking Long-Range Memory for Causal Long Video Generation},
  author={Chen, Yuzhuo and Shi, Huafeng and Wang, Xinyu and Wang, Yucheng and Hong, Haoqin and Zhang, Guoxin and Ma, Zehua},
  year={2026}
}

When using LSMDC titles, also cite Rohrbach et al., IJCV 2017.